Assessing the Condition of Children

Nicholas Zill
Mary Jo Coiro

Child Trends, Inc.

  • Proposed child welfare legislation challenges states to initiate new programs to help children who have been neglected, abused, or abandoned by their parents. This paper describes ways to appraise the condition of children and the functioning of families in order to evaluate the effects of child welfare reform efforts. The assessment approach grows out of a "whole child" perspective, encompassing physical health, cognitive development, emotional well-being, and moral development and social behavior. Specific measures to assess these aspects of child development are described, most of which have shown meaningful differences between groups in large-scale studies. However, child welfare situations present special challenges: the difficulty of reliably assessing the developmental status of young children; applicability of measures to low-income and minority children and very dysfunctional families; and the problem of obtaining accurate, unbiased information from parents. Several research strategies are proposed to combat these difficulties. One promising development for the future of child welfare evaluations is that a number of these strategies have been incorporated into large-scale, longitudinal evaluation studies of welfare-to-work, child care, and integrated service delivery programs.
  • The family is the most important institution in children's lives. Our society relies on families to perform functions that are critical to the survival and development of children, especially young children. Among these functions are: providing physical necessities, like food, clothing, and shelter; protecting children from harm and supervising their daily activities; giving affection, praise, and other forms of emotional support; and applying firm but not harsh discipline when it is required (Baumrind, 1971; Horowitz, 1989).

    We expect children to form an intense, irreplaceable bond with their parents, a bond that seems to be critical for normal social development in the human organism (Bowlby, 1969; Rutter, 1981). This bond helps to nurture and shape the child's developing sense of self; to steer social behavior into acceptable channels (Erikson, Sroufe & Egeland, 1985); and to motivate accomplishments that will be gratifying to the parents (Bretherton & Waters, 1985; Egeland & Farber, 1984).

    We count on parents to serve as the child's first instructors, and to continue intellectual stimulation and encouragement for learning after the child has started school. We presume that parents will teach children right from wrong, and respect for the rights of others, by both precept and example. We also expect that family members will pass on the traditions and values of the political, cultural, and religious communities of which the family is a part.

    When families fail to perform these functions, or perform them badly, everyone pays a price. Youngsters may be injured or even die, be delayed in their development, or develop abnormally. Schools may have to compensate for a lack of intellectual stimulation at home or be forced to deal with conduct problems that have their roots in family turmoil. In extreme cases, public agencies or private charities may be required to take over family functions. Rarely do these organizations carry out child rearing tasks with the dedication and efficiency that a parent who loves the child usually brings to them.

    Unfortunately, many families today are unable to fulfill their traditional function, and are instead neglecting, abusing, or abandoning their children. More than 2 million reports of child maltreatment are received by child protection agencies across the U.S. each year (U.S. House of Representatives Select Committee on Children, 1989, pp. 68-69, 190-191). State and local governments have had to become increasingly involved with these families, often by removing children from the home and placing them in foster homes. More than 400,000 American children are in substitute care at any given time (Tatara, 1991). However, children do not always fare better with "substitute families;" furthermore, the caseloads of family service agencies are critically overburdened (U.S. House of Representatives Select Committee on Children, 1987). Thus, efforts are underway to help families remain intact and well-functioning and to explore alternative approaches to substitute care placement. The most promising programs appear to be those that emphasize preventive services such as parenting classes and respite care, rather than crisis intervention when the family environment has already deteriorated significantly.

    Proposed child welfare legislation contemplates a process whereby the states initiate demonstration projects to test whether new approaches and programs for families in crisis can do a better job than existing procedures and services of helping children at risk. The ultimate test of whether these programs work should not be merely some administrative criterion, such as a reduction in the number of foster care placements that the state is required to make, but a measurable improvement in the condition of the children themselves. But how does one discern whether children are better off than they were under earlier programs? How does one measure the condition of children? That question is the focus of this paper.

    Note that the kind of information researchers need to evaluate reform efforts does not have to be as comprehensive nor as scrupulously accurate as the information required for placement decisions involving individual children. Rather, program analysts need reasonably reliable and valid measures that can be administered to two or more groups of children to see whether the development, well-being, or life circumstances of one group is demonstrably better than that of another, comparison group.

    It would be helpful to have some sense of how significant a given difference is for the later development and life prospects of children in the two groups. It would be even more helpful to be able to attach a dollar figure to the difference; to be able to say, for example, that an observed difference in average vocabulary test scores among 4-year-old children would result in a savings to the state of "X" dollars in future expenditures for remedial instruction, grade repetition, special educational services, or teen pregnancies. One of the reasons that the longitudinal study of the Perry Preschool Program has had such an impact in the educational arena is that that study has been able to show the long-term differences between those who received the program and those who did not, and to attach dollar figures to the differences (Berrueta-Clement, Schweinhart, Barnett, Epstein, & Weikart, 1984). Unfortunately, the knowledge base that would permit developmental forecasts and meaningful cost-benefit analyses is often lacking, short of waiting for the results of new 10- to 20-year longitudinal studies. But it is also true that better use could be made of knowledge that is already available.

    It is important that we consider carefully what the major indicators of child and family well-being are, and what outcomes might be associated with the effects of new programs. Because child developmentalists have already developed a wide range of valid, reliable instruments and procedures for assessing these indicators, it should be possible to build on this knowledge, while applying it to child welfare demonstration projects.

    Indicators of Children's Development and Well-Being

    Most child development experts do not believe it possible or desirable to summarize the developmental status and well-being of a child or a group of children with a single number, be it a health index, a percentile on a growth chart or developmental scale, an IQ test score, or whatever. Child experts are more comfortable with a developmental profile, covering several different dimensions or domains (Sattler, 1982). A list of these domains would include:

  • ´ physical health, nutrition, and safety;

    ´ cognitive development and academic achievement;

    ´ emotional well-being; and

    ´ moral development and social behavior.

  • The following sections illustrate the types of data that make up each of these domains (see also Table 1).

    Physical Health, Nutrition, and Safety

    Examples of basic health indicators are whether the child's height is within the normal range for his or her age, whether weight is appropriate for height, whether there are any obvious signs of malnutrition, and whether there have been any significant delays in growth or motor development. Pediatricians would also want to find out whether the child has any life threatening or life-shortening diseases, or any chronic illness or impairment that causes discomfort or limits play or learning.

    Because non-intentional injury and violence are important parts of the current child health picture (Gallagher, Finison, Guyer, & Goodenough, 1984), even in non-abusive families, doctors would try to find out whether the child has a history of frequent visits to hospital emergency rooms or has had numerous untreated injuries in the recent past. They would also inquire about the presence of dangerous conditions in the child's daily environment, and, if possible, examine the child's body for the presence of questionable bruises, scars, or broken bones.

    Questions about access to medical care are not direct measures of the child's current condition, but they do provide information that is relevant to future health. Important to know here are whether the child has a regular source of medical and dental care, when he or she last received a check-up, whether he or she has been screened for possible vision and hearing impairments, and whether he or she has received appropriate immunizations.

    Under normal circumstances, one of the most robust and predictive indicators of a child's physical condition is a simple rating of the child's general health by the mother or other knowledgeable care giver, using a five-category scale ranging from "excellent" to "poor" (Eisen, Donald, Ware, & Brook, 1980). However, in child welfare cases, circumstances are often not normal, and there may be reason to doubt the trustworthiness of health ratings from the mother. The broader question of the validity of parental reporting in child welfare cases is considered in greater detail later in this paper.


    Table 1
    Possible Measures


    Developmental Domains Examples of Existing Measures

    Physical Health and Safety -National Health Interview Survey on Child Health

    -Measurement of child's height and weight

    Cognitive Development -Early Screening Inventory (ESI)

    and Academic Achievement -Peabody Picture Vocabulary Test (PPVT)

    -Woodcock-Johnson Scales

    -Grade Repetition

    -Special education or remedial placement

    Emotional Well-Being -Personal Maturity Scale (Teacher report)

    and Problem Behavior -Behavior Problems Index (Parent Report)

    -School conference for problem behavior

    Social Development -National Survey of Children scale on

    relationship with parents

    -Harter and Pike Teacher Rating scale of

    competence and social acceptance

    -Classroom sociometrics_

    Family Functions Examples of Existing Measures

    Fulfillment of Physical Needs -Dunst Family Resources Scale

    (Food, Clothing, & Shelter)

    Protection from Harm, -National Commission on Children Survey of

    Monitoring and Supervision Parents and Children

    Intense, Irreplaceable Bond -Ainsworth Strange Situation

    -Waters and Deane Attachment Q-Sort

    -Parental Acceptance-Rejection scale

    Emotional Support -HOME scale

    -Feshbach Parental Empathy scale

    Intellectual Stimulation -HOME scale

    -National Household Education Survey

    Aspirations and Expectations -National Survey of Children Educational

    Aspirations Scale

    -Child Development Knowledge Scales


    Cognitive Development and Academic Achievement

    This heading encompasses the child's attainment of the skills, knowledge, concepts, and strategies that are needed to succeed in school and, eventually, to deal with the increasingly complex worlds of work, consumption, family life, and citizenship. A wide variety of tests have been designed to measure various aptitudes and achievements among children of different ages (see Sattler, 1982, on Stanford-Binet Intelligence Scale, Wechsler Intelligence Scale for Children, McCarthy Scales of Children's Abilities, and Peabody Picture Vocabulary Test).

    When the child reaches school age, information can be obtained about the child's standing in class, progress through the grades, and receipt of remedial instruction or special educational services. Also useful are questions about educational aspirations and expectations: how far the parent would like the child to go in school, and how far the parent thinks the child really will go, the way it looks now. Similar questions can be put to children themselves, once they reach the later grades of elementary school.

    Emotional Well-Being

    The domain of emotional well-being covers the child's overt behavior and his or her moods and feelings, as they can be inferred from parent or teacher reports (Achenbach & Edelbrock, 1981), through direct observation of children's facial and vocal expressions and deportment, or through interviews with older children and adolescents. Well-being implies freedom from debilitating emotional states, such as chronic fear or depression, and the absence of detrimental behavior patterns, such as hyperactive or withdrawn behavior. It also signifies the presence of positive emotions, such as interest and enjoyment, and constructive activities, such as exploration and play.

    Common patterns of problem behavior in childhood are: aggressive-antisocial behavior (cheating, lying, bullying other children, vandalism, stealing); phobias and anxiety disorders; childhood depression (appearing unhappy much of the time, not getting involved with other children, not participating in play or school activities); immature dependency (being extremely clingy, dependent, and demanding, even with unfamiliar adults); and the attention-deficit hyperactive disorder (the child cannot sit still or concentrate for any length of time and is extremely restless and overactive) (Achenbach, 1982; Tuma, 1989).

    Moral Development and Social Behavior

    Under this heading comes information about whether the child has a strong, secure relationship with at least one parent or parent-substitute. This means that the child feels close to that parent, enjoys doing things or just being with the parent, feels loved by the parent, wants to be like the parent when he or she grows up, and cares what the parent thinks about his or her behavior and achievements. Also included here is information about how the child gets along with peers and playmates, whether he or she is developing age-appropriate social competencies like sharing and cooperative play, and whether he or she feels liked and respected by other children (Harter, 1982; Laosa, 1979). Subsumed here as well are indicators of the young person's development of culturally-appropriate values, standards, and attitudes, such as empathy and respect for the rights of others (Wright, 1971).

    Measures Used in Large-Scale Studies of Children

    There are existing tests, questionnaire items, and rating scales that can be used to assess different aspects of child development and well-being in a standardized, valid, reliable, and reasonably succinct fashion. Some of the measures have been successfully used in studies of national probability samples of families with children, as in the National Survey of Children (Peterson & Zill, 1986; Allison & Furstenburg, 1989), the Child Supplement to the National Longitudinal Survey of the Labor Market Experience of Youth (Baker & Mott, 1989), the National Health Interview Survey on Child Health (National Center for Health Statistics, 1989; Dawson, 1991; Zill, Moore, Smith, Stief, & Coiro, 1991); and the national survey conducted for the National Commission on Children (National Commission on Children, 1991). Examples of such measures are:

  • ´ questions on children's health status from the Rand Health Insurance Experiment (Eisen, Donald, Ware, & Brook, 1980) and the National Health Interview Survey on Child Health (NHIS-CH);

    ´ questions on children's access to and receipt of preventive care from the NHIS-CH;

    ´ the Peabody Picture Vocabulary Test and other cognitive tests used in the Child Supplement to the National Longitudinal Study of Youth (NLSY-CS);

    ´ questions on grade repetition and standing in class used in the National Survey of Children (NSC) and the NHIS-CH;

    ´ the Behavior Problems Index (Zill, 1990), a short scale for parental rating of children's problem behavior that was used in the NSC, the NHIS-CH, and the NLSY-CS;

    ´ the Personal Maturity Scale, a short scale for teacher rating of children's behavior and adjustment in the classroom, used in the NSC and in a longitudinal study of pupils from low-income families in Baltimore (Alexander & Entwisle, 1988);

    ´ questions on children's need for and receipt of psychological help for emotional and behavior problems, used in the NSC and the NHIS-CH (Zill & Schoenborn, 1990); and

    ´ questions to adolescent children, on the quality of their relationships with their mothers and fathers, from the NSC.

  • Measures such as these have been used to show developmental differences between children in intact two-parent families and those in single-parent families (Allison & Furstenburg, 1989; Peterson & Zill, 1986; ) or step-families (Dawson, 1991; Zill, 1988; Zill & Schoenborn, 1990); between children adopted in the first year of life and those adopted later (Zill, 1985); and between welfare children and non-poor children (Zill et al., 1991). There are several potential problems in using these measures in child welfare evaluations, however. These are discussed below._

    Indicators of the Family Situation

    With respect to the kinds of information one would want to collect about the family situations of children in child welfare cases, a useful set of categories is obtained by listing the things that parents are supposed to do for children in our society. As described at the beginning of this paper, the functions of families with children include:

  • ´ fulfilling children's physical needs;

    ´ protecting children from harm, and monitoring and supervising their activities;

    ´ forming an intense, irreplaceable bond between child and parent (or parent substitute);

    ´ providing emotional support and firm but not harsh discipline;

    ´ furnishing intellectual stimulation;

    ´ having high but not unreasonable aspirations and expectations for children's conduct and achievement;

    ´ providing moral guidance and example; and

    ´ transmitting family traditions and the broader cultural heritage to children.

  • Through questions to parents, teachers, and others (such as knowledgeable relatives or family friends), and through direct observation of the home and of family members interacting with one another, one can determine whether and how well the family or foster family is carrying out these functions. If parents or guardians do none of these things for their children, or very few of them, then one could argue that there really is no family worth preserving. More frequently, however, the dilemma in neglect and abuse cases is that parents fulfill some of these functions but are seriously deficient in others.

    Existing measures of the home environment

    There are existing, validated instruments that can be used to assess several of the major functions of families. For example, the Home Observation for Measurement of the Environment (HOME) Scale is a well-validated and widely-used instrument developed by Robert Bradley and Bettye Caldwell. The HOME scale is designed to appraise whether the child's home is an environment that nurtures the child's intellectual and emotional development and helps to prepare him or her for the challenges of school (Bradley & Caldwell, 1981; Caldwell & Bradley, 1984). It assesses the orderliness, cleanliness, and safety of the physical environment, the regularity and structure of the family's daily routine, the amount of intellectual stimulation available to the child, and the degree of emotional support provided by the parents. It does this with a combination of questions asked of the parent and items to be completed by the interviewer after spending time in the home observing the child's physical surroundings and the parent and child interacting with one another.

    Abbreviated versions of the HOME scale were developed especially for the NLSY-CS, with different forms being used for infants and toddlers, preschoolers, and elementary school-aged children (Baker & Mott, 1989). The 12 parent report items and 10 interviewer observations that comprise the abbreviated scale for families with children aged 3-5 are shown in Table 2.

    The NLSY HOME scale proved to have reasonable reliability, with the total score having an alpha reliability of .70 for preschool-aged children. However, the subscale measuring "emotional support" (alpha reliability = .49 for children aged 3-6) was less reliable that the subscale that measured "intellectual stimulation" (alpha reliability = .69) (Baker & Mott, 1989)._

    _ In a study that made use of data from the NLSY-CS, significant relationships between welfare and poverty status and HOME scores were found for black, Hispanic, and non-minority children (Zill et al., 1991). (See Tables 3 and 4.) Only about one-third of preschool children from AFDC families received intellectual stimulation and emotional support from their parents comparable to that received by most children in families that were neither poor nor welfare dependent. Preschoolers in poor, non-welfare families also tended to have home environments that were less than optimal in terms of support for emotional health and school achievement. These differences were found among black, Hispanic, and non-minority children. However, AFDC children in black and Hispanic families were generally more disadvantaged than those in non-minority welfare families.


    Table 2
    Contents of Abbreviated HOME Scale (NLSY, Children Aged 3-5)


    Parent Report Interviewer Observation

    ´Child read to several times per week ´Parent's voice conveyed positive feeling ´Child has 10 or more books of own about child

    ´Family gets at least one magazine ´Parent conversed with child at least once

    ´Child has use of record or tape player during visit

    and tapes of own ´Parent caressed, kissed, or hugged child at

    ´Parent has helped to teach child numbers, least once

    alphabet, colors, shapes and sizes ´Parent introduced interviewer to child by

    ´Child has some choice in foods to eat name

    ´Parent limits hours of television ´Parent did not physically restrict, shake, or

    ´Parent does not hit back when child hits grab child

    ´Child taken on outings at least monthly ´Parent did not slap or spank child during

    ´Child eats meal with both mother and visit

    father figure once a day or more ´Child's play environment appears safe

    ´Child taken to museums at least yearly ´Interior of home not dark or perceptually

    ´Child spanked less than twice in one week monotonous

    ´All visible rooms reasonably clean

    ´All visible rooms minimally cluttered


    The same study made use of data from the NHIS-CH on health-related conditions in the home. Conditions in many AFDC and non-welfare poor families proved to be less healthful than those in non-poor families with respect to such indicators as parental smoking, children's use of seat belts, and children having regular and reasonable bedtimes.

    As shown by the relatively low reliability of the "emotional support" subscale of the HOME scale, our current ability to assess family environments is far from ideal. The HOME scale has also been accused of being biased against children in single-parent families and of embodying middle-class child rearing values. Nevertheless, HOME scores have been found to be predictive of school performance among minority as well as non-minority children (Bradley & Caldwell, 1981). Also, the abbreviated HOME scale has been found to relate to children's achievement, correlating with vocabulary, reading, and math tests given in the Child Supplement (Dubow & Luster, 1990; Menaghan & Parcel, 1991; Morrison, Myers, & Winglee, 1990; Parcel & Menaghan, 1989). Significant correlations remained even when family social and economic status and mother's scores on the Armed Forces Qualifying Test were controlled (Moore & Snyder, 1991).


    Table 3
    The Quality of Children's Home Environments (HOME Scale Scores) by Ethnicity and Welfare and Poverty Status of Their Families, Children Aged 3-5 Born to Mothers Aged 14-25 at Birth of Child, United States, 1986.


    Proportion of Black Children Whose Home Environments Were:

    Quality of Home

    Environment Deficient Below Average Supportive

    (Total HOME Score)1 (<15) (15-18) (19+) TOTAL

    All children aged 3-5 in NLSY-CS 24% 42% 34% 100%

    Welfare/Poverty Status:

    In AFDC family 38% 43% 19% 100%

    In poor, non-AFDC family 32% 48% 20% 100%

    In non-poor, non-AFDC family 14% 39% 47% 100%

    (contingency coefficient) (.32***)

    Proportion of Hispanic Children Whose Home Environments Were:

    Quality of Home

    Environment Deficient Below Average Supportive

    (Total HOME Score) (<15) (15-18) (19+) TOTAL

    All children aged 3-5 in NLSY-CS 21% 37% 42% 100%

    Welfare/Poverty Status:

    In AFDC family 31% 44% 25% 100%

    In poor, non-AFDC family 38% 44% 18% 100%

    In non-poor, non-AFDC family 14% 34% 52% 100%

    (contingency coefficient) (.30***)

    Proportion of Non-Minority Children Whose Home Environments Were:

    Quality of Home

    Environment Deficient Below Average Supportive

    (Total HOME Score) (<15) (15-18) (19+) TOTAL

    All children aged 3-5 in NLSY-CS 7% 25% 68% 100%

    Welfare/Poverty Status:

    In AFDC family 12% 40% 48% 100%

    In poor, non-AFDC family 18% 37% 45% 100%

    In non-poor, non-AFDC family 4% 22% 74% 100%

    (contingency coefficient) (.25***)


    *** p < .001

    1Total HOME Scale scores ranged from zero to 22.

    Source: Zill, N., Moore, K.A., Smith, E.W., Stief, T., & Coiro, M.J. (1991).


    Table 4
    Reading to Child, Book Ownership, and Television Watching by Ethnicity and Welfare and Poverty Status of Family, Children Aged 3-5 Born to Mothers Aged 14-25 at Birth of Child, United States, 1986.


    Proportion of Children for Whom Statement Applies, by Ethnicity:

    HOME Scale Items All

    Ethnic Non-

    Mother reads stories to child Groups Black Hispanic Minority

    three or more times a week

    Welfare/Poverty Status:

    In AFDC family 42% 30% 33% 53%

    In poor, non-AFDC family 36% 27% 28% 42%

    In non-poor, non-AFDC family 57% 37% 43% 61%

    (contingency coefficient) (.19***) (.17***) (.18***) (.16***)

    Child has 10 or more books

    Welfare/Poverty Status:

    In AFDC family __ 51% 28% 39% 72%

    In poor, non-AFDC family 59% 33% 23% 78%

    In non-poor, non-AFDC family 81% 54% 57% 88%

    (contingency coefficient) (.29***) (.29***) (.27***) (.22***)

    Television is on in home

    7 or more hours every day

    Welfare/Poverty Status:

    In AFDC family 55% 60% 47% 53%

    In poor, non-AFDC family 44% 44% 42% 44%

    In non-poor, non-AFDC family 35% 43% 31% 34%

    (contingency coefficient) (.20***) (.19***) (.18***) (.16***)


    *** p < .001

    Source: Zill, N., Moore, K.A., Smith, E.W., Stief, T., & Coiro, M.J. (1991).


    Special Challenges Posed by Child Welfare Assessments

    Attempts to apply the assessment procedures described above to children and families involved with child welfare agencies face special difficulties. There are several reasons for this:

  • ´ A disproportionate number of neglected or abused children are young, that is, of preschool age (U.S. House of Representatives Select Committee on Children, 1989, p. 191). Assessing developmental status is more challenging with preschool than with school-aged children (Sattler, 1982).

    ´ A disproportionate number of neglected and abused children come from low-education black or Hispanic families. Because much of the child development literature is based on work with children from middle-class white families, instruments and norms drawn from this literature may not be suitable for child welfare populations.

    ´ Assessments of child health and well-being are heavily dependent on information provided by the mother or another caregiver who is well familiar with the child's history and customary behavior. In child welfare cases, such an informant may be unavailable or the trustworthiness of parental reports may be questioned because of substance abuse, mental illness, or motivation to portray family conditions in a favorable light.

    ´ By definition, the family situations that come to the attention of child welfare agencies are those in which living conditions are extreme. Indicators that discriminate well within the normal range of variation may be less informative or even useless with families at the outer edge of the distribution.

  • Research Strategies for Improving Assessments

    There are research strategies that could be adopted to deal with these difficulties. One is to apply a range of child and family measures to substantial samples of the child welfare caseload, as well as to samples of disadvantaged families who show no evidence of child neglect or abuse. This would provide normative data on children from low-income families as well as evidence as to which measures "work" in differentiating between neglectful and non-neglectful families. The measures used should be selected to cover a wide range of circumstances, including basic child survival indicators like measures of malnutrition and growth retardation. Normally, such "Third World" indicators show little variation in population samples of children from the U.S. or other developed nations. They may prove useful, however, in describing the extreme circumstances with which child welfare agencies must deal.

    A second strategy is to explore the use of new methods for gathering data about families at risk. One such method might be the use of third parties as informants. Attempts should be made to locate relatives or friends of the family who are relatively familiar with the child and parents. These third parties should be interviewed with specially designed questionnaires that tap some of the same areas that would normally be covered in interviews with the mother or mother substitute. The reliability and validity of the information so obtained should then be carefully evaluated. Although this source of information would not be available in all cases, it could prove to be a useful adjunct to existing sources.

    A method that shows great promise in child welfare applications is direct observation of the child and mother (or parent-substitute) interacting with one another in a series of structured activities and tasks. The tasks are selected so as to be appropriate to the child's age and the family processes that are to be assessed. The interaction is videotaped and subsequently coded by trained observers who rate such things as the quality of the mother-child interaction, the apparent emotional state of mother and child, the types of disciplinary methods and teaching tactics used by the mother, and so on. The same tape can be coded by several different raters and from several different substantive perspectives. Such structured observation methods have been used by child development researchers in laboratory studies, and predict children's later academic achievement and behavioral and emotional adjustment (Egeland, 1991). However, the application of these observational measures to large-scale, representative samples and less-than-ideal family situations has only begun to be investigated.

    In order to obtain more stable and definitive assessments of the effects of different family preservation or substitute care programs on the development of young children, it may be necessary to extend these studies longitudinally. In other words, the evaluations would continue until the children are at least of late elementary-school ages. Then it would be possible to look at indicators such as standing in class, absenteeism, grade repetition, use of special educational resources, conduct and disciplinary problems, and children's reports on the quality of their relationships with parents or parent substitutes. These measures are "proven" in the sense of being more predictive than assessments in the preschool years of later developmental problems of great practical concern, such as school dropout, delinquency, teen pregnancy, and joblessness. Unfortunately, such long-term panel studies are costly in both time and money, pose problems of sample attrition, and mean that the answer to the question of how well a program works is long delayed.

    Important Parallel Evaluations

    One promising development for the future of child welfare evaluations is that a number of the methods and strategies described above are currently being used or are about to be used in large-scale, longitudinal evaluation studies of welfare-to-work, child-care, and integrated services delivery programs (Smith, 1991). Examples are: the Teen Parent Demonstration Project being carried out by Mathematica Policy Research, Inc. with support from the Rockefeller Foundation and the U.S. Department of Health and Human Services; the evaluation of the Comprehensive Child Development Centers being carried out by Abt Associates for the Office of Human Development Services, DHHS; and the JOBS Evaluation and New Chance study being conducted by the Manpower Demonstration Research Corporation, Child Trends, and Humanalysis, Inc., with support from, in the first instance, the Assistant Secretary for Planning and Evaluation and the Family Support Administration of DHHS and, in the second, a number of private foundations.

    For the potential applications discussed here, there are two encouraging things to report about these new evaluation studies that include measures of children's development. First, they are dealing with populations that are quite similar to, though perhaps not as extremely troubled as, the families with which child welfare agencies must deal. That means that these groups are developing a base of assessment experience in non-white, non-middle class families. They will get a better sense of what works and what does not under these conditions.

    Second, the different groups of investigators are actually communicating with one another and sharing information and experience. Thanks to the good offices of people like Barbara Blum and Sheila Smith at the Foundation for Child Development and Shelby Miller at the Ford Foundation, evaluation researchers have been meeting regularly as part of something called the Two-Generation Research Network. Inasmuch as many of the same organizations may well be conducting the child welfare reform experiments that are being discussed in this symposium, these proposed experiments stand to gain by the growing body of collective experience, and, one hopes, methodological wisdom.

    All of which is not to say that there will not be continuing problems with the application of measures of child development and family functioning to families in crisis. The hope, though, is that developmental science can eventually be of significant assistance to those children who are at greatest risk in our society.


    Achenbach, T. (1982). Developmental psychopathology (second ed.). New York: John Wiley & Sons.

    Achenbach, T., & Edelbrock, C. (1981). Behavioral problems and competencies reported by parents of normal and disturbed children aged four through sixteen. Monographs of the Society for Research in Child Development, 46(Serial No. 188).

    Alexander, K.L., & Entwisle, D.R. (1988). Achievement in the first two years of school: Patterns and processes. Monographs of the Society for Research in Child Development, 53 (Serial No. 218).

    Allison, P.D., & Furstenburg, F.F., Jr. (1989). How marital dissolution affects children: Variations by age and sex. Developmental Psychology, 25, 540-549.

    Baker, P.C., & Mott, F.L. (1989). A guide and resource document for the National Longitudinal Survey of Youth 1986 child data. Columbus: The Ohio State University, Center for Human Resource Research.

    Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology Monographs, 4, 1-103.

    Berrueta-Clement, J., Schweinhart, L.J., Barnett, W.S., Epstein, A.S., & Weikart, D.P. (1984). Changed lives: The effects of the Perry Preschool Program on youth through age 19. In Monographs of the High/Scope Educational Research Foundation (No. 8). Ypsilanti, MI: High/Scope Press.

    Bowlby, J. (1969). Attachment and loss: I. Attachment. London: Hogarth Press.

    Bradley, R.H., & Caldwell, B.M. (1981). The HOME inventory: A validation of the preschool scale for black children. Child Development, 52, 708-710.

    Bretherton, I., & Waters, E. (Eds.). (1985). Growing points in attachment theory. Monographs of the Society for Research in Child Development, 50 (Serial No. 209).

    Caldwell, B.M., & Bradley, R.H. (1984). Home observation for the measurement of the environment. Little Rock: University of Arkansas.

    Dawson, D. (1991). Family structure and children's health and well-being: Data from the 1988 national health interview survey on child health. Journal of Marriage and the Family, 53, 573-584.

    Dubow, E.F., & Luster, T. (1990). Adjustment of children born to teenage mothers: The contribution of risk and protective factors. Journal of Marriage and the Family, 52, 393-404.

    Egeland, B. (1991). Unpublished data showing correlations between mother and child rating scales and child outcomes. University of Minnesota, Minneapolis.

    Egeland, B., & Farber, E.A. (1984). Infant-mother attachment: Factors related to its development and changes over time. Child Development, 55, 753-771

    Eisen, M., Donald, C.A., Ware, J.E., Jr., & Brook, R.H. (1980, May). Conceptualization and measurement of health for children in the health insurance study. Santa Monica, CA: Rand.

    Erikson, M.F., Sroufe, L.A., & Egeland, B. (1985). The relationship between quality of attachment and behavior problems in preschool in a high-risk sample. Monographs of the Society for Research in Child Development, 50 (Serial No. 209), 147-156.

    Gallagher, S.S., Finison, K., Guyer, B., & Goodenough, S. (1984). The incidence of injuries among 87,000 Massachusetts children and adolescents: Results of the 1980-81 statewide childhood injury prevention program surveillance system. American Journal of Public Health, 74, 1340-1347.

    Harter, S. (1982). The perceived competence scale for children. Child Development, 53, 87-97.

    Horowitz, F.D.(Ed.). (1989). Children and their development: Knowledge base, research agenda, and social policy application. American Psychologist, 44 (Special issue), 95-445 .

    Laosa, L.M. (1979). Social competence in childhood: Toward a developmental, socioculturally relativistic paradigm. In Ken, M.W., & Rolf, J.E. (Eds.), Primary prevention of psychopathology. Vol. III: Social competence in children (pp. 253-279). Hanover, NH: University Press of New England.

    Menaghan, E.G., & Parcel, T.L. (1991). Determining children's home environments: The impact of maternal characteristics and current occupational and family conditions. Journal of Marriage and the Family, 53, 417-431.

    Morrison, D.R., Myers, D.E., & Winglee, M. (1990). The effects of maternal work and child care during the first three years of life on children's cognitive abilities. Paper presented at the Meetings of the Population Association of America, Toronto, Canada.

    Moore, K.A., & Snyder, N.O. (1991). Cognitive attainment among firstborn children of adolescent mothers. American Sociological Review, 56, 612-624.

    National Center for Health Statistics. (1989). Vital statistics of the United States, 1987, Vol. 1, Natality. Washington, DC: GPO, DHHS Publication No(PHS) 89-1100.

    National Commission on Children. (1991). Speaking of kids: A national survey of children and parents. Washington, DC: Author.

    Parcel, T.L., & Menaghan, E.G. (1989). Child home environment as a mediating construct between SES and child outcomes. Draft manuscript, Ohio State University, Department of Sociology, Columbus.

    Peterson, J.L., & Zill, N. (1986). Marital disruption, parent-child relationships, and behavioral problems in children. Journal of Marriage and the Family, 48, 295-307.

    Rutter, M. (1981). Maternal deprivation reassessed (2nd ed.). Great Britain: Penguin Books.

    Sattler, J. (1982). Assessment of children (3rd ed.). San Diego, CA: Author.

    Smith, S. (1991). Two-generation program models: A new intervention strategy. Society for Research in Child Development. Social Policy Report, 5(1), 1-15.

    Tatara, T. (1991, November). Some additional explanations for the recent rise in the U.S. child substitute care population. VCIS Research Notes (No. 5, November). Washington, DC: American Public Welfare Association.

    Tuma, J.M. (1989). Mental health services for children. American Psychologist, 44, 188-199.

    U.S. House of Representatives, Select Committee on Children Youth and Families. (1987). Abused children in America: Victims of official neglect. Washington, DC: U.S. Government Printing Office.

    U.S. House of Representatives, Select Committee on Children Youth and Families. (1989). U.S. children and their families: Current conditions and recent trends, 1989 (3rd ed.) (Prepared by Child Trends). Washington, DC: U.S. Government Printing Office.

    Wright, D. (1971). The psychology of moral behavior. Middlesex, England: Penguin Books.

    Zill, N. (1985, April). Behavior and learning problems among adoptive children: Findings from a national survey of children. Tables presented at the Meetings of the Society for Research in Child Development, Toronto. Washington, DC: Child Trends.

    Zill, N. (1988). Behavior, achievement, and health problems among children in stepfamilies: Findings from a national survey of child health. In Hetherington, E.M., & Arasteh, J. (Eds.). The impact of divorce, single parenting, and step-parenting on children (pp. 325-368). Hillsdale, NJ: Lawrence Erlbaum.

    Zill, N. (1990). The Behavior Problems Index [descriptive material]. Washington, DC: Child Trends.

    Zill, N., Moore, K.A., Smith, E.W., Stief, T., & Coiro, M.J. (1991). Life circumstances and development of children in welfare families: A profile based on national survey data. Washington, DC: Child Trends.

    Zill, N., & Schoenborn, C.A. (1990, November 16). Health of our nation's children: Developmental, learning, and emotional problems, United States, 1988. Advance Data from Vital and Health Statistics, No. 190. Washington, D.C.: U.S. Department of Health and Human Services.